# How to solve a Poisson equation with only Neumann boundaries?

 Register Blogs Members List Search Today's Posts Mark Forums Read

 March 31, 2009, 13:14 How to solve a Poisson equation with only Neumann boundaries? #1 New Member   Join Date: Mar 2009 Posts: 7 Rep Power: 9 I need to solve a pressure Poisson equation with only Nuemann boundaries with F.D. method. How we can get the unique results without introducing any errors into? Many thanks in advance!

 March 31, 2009, 20:46 #2 Member   Join Date: Mar 2009 Posts: 33 Rep Power: 9 Specify a pressure value (any value) at just one point (anywhere) on the boundary.

 April 1, 2009, 03:23 #3 New Member   Join Date: Mar 2009 Posts: 7 Rep Power: 9 But if we set up a reference point at the boundary, certain error will introduced. any possibility we do not do that? Cheers!

 April 1, 2009, 09:56 #4 Member   Join Date: Mar 2009 Posts: 33 Rep Power: 9 > certain error will introduced. What kind of error? There is no unique exact solution to this problem, not just for discretely but analytically: the solution is unique only up to an additive constant.

 April 1, 2009, 11:12 #5 New Member   Join Date: Mar 2009 Posts: 7 Rep Power: 9 yes, that is the thing. If we set one of the point on the boundary as reference point, which is Dirichlet bounday. But how we can make sure it is also satisfying the Nuemann boundary condition? If it is not, error on the boundary will be introduced. Cheers!

April 2, 2009, 02:56
#6
Senior Member

Rami Ben-Zvi
Join Date: Mar 2009
Posts: 148
Rep Power: 9
Quote:
 Originally Posted by mbgnfrx2 yes, that is the thing. If we set one of the point on the boundary as reference point, which is Dirichlet bounday. But how we can make sure it is also satisfying the Nuemann boundary condition? If it is not, error on the boundary will be introduced. Cheers!
You may prescribe the reference pressure value at any point, not necessarily on the boundary.

April 2, 2009, 04:01
#7
Member

MrFluent
Join Date: Mar 2009
Posts: 33
Rep Power: 9
Quote:
 Originally Posted by mbgnfrx2 I need to solve a pressure Poisson equation with only Nuemann boundaries with F.D. method. How we can get the unique results without introducing any errors into? Many thanks in advance!
When you construct your matrix A.
Make Albnd = 0, and do not add its contribution to Ap (or diagonal element).

Imagine that your Ap (diagonal element) is like this

Ap = -Sum (Al)

To get Ap you usually do Ap = Ap - Albnd
Src = Src + albnd * phi (if phi was fixed value).

Instead now you do not modify Src. Do not modify Ap, just make Albnd = 0.

solve matrix.

You get some solution, pick a point from these Phi you just calculated. Lets say iref.

phi_ref = Phi[ iref ] ;

For all
{
Phi = Phi_from_matrix - phi_ref.
}

Thats all. Painless and simple.

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post Allan Main CFD Forum 11 March 17, 2011 10:39 Beginner Main CFD Forum 0 September 6, 2007 09:18 cregeo Main CFD Forum 8 July 26, 2004 04:58 windhair CFX 3 January 27, 2004 22:09 Linfeng BI Main CFD Forum 4 November 21, 2002 00:46

All times are GMT -4. The time now is 20:15.